Cube Troubleshooting Best Practice

[gweisbecker]

Hi,

just now i'am in the implemantation of a new cucm cluster for one of my customers. I have 4 CUCM's in the Cluster,

and a centralized SIP Trunk with 2 Cube's ( 2 x 3925 )  in active/active mode. Every Cube should handle a maximum of

500 concurrent calls at once, which will not be problem for the Performance of the ISR's.

What I'am thinking about is how to Troubleshoot such a Cube, if there are 500 Calls at once, and I have to collect Traces like

the  "SIP Call Message tracing" if neccessary. I think this will result in the same as if I would enable a" Debug IP Packet all" Trace on a Router.

What is the best Practice to Monitor and Troubleshoot such a big Cube ?

Regards

Gerhard

[Rob Huffman]

Hi Gerhard,

You have a unique opportunity right now to engage

Edson from Cisco in this "Ask the Expert" event

Ask the Experts: Configuring and Troubleshooting Cisco SIP CUBE/Gateways and MGCP Gateways

https://supportforums.cisco.com/thread/2209795?tstart=0

Cheers!

Rob

"Talk about a dream
Try to make it real" 

- Springsteen

[Ayodeji Okanlawon]

Gerhard,

In scenarios or enviroment like this one tool that is available to you is the feature called Per call Debugging on the CUBE..

CUBE Per-Call Debugging (PCD)

• Useful for CUBE under high call volume
  
• Available on all CUBE(Ent) ASR releases and in 15.1(2)T and later on ISR
  
• All the debug pertaining to a particular call goes into a buffer
  
• “Trigger-points” looks for specific info in the buffers to export the debug info to an output destination
  
• Can trigger based on user-defined criteria or log every call
  

          SIP 4XX, 5XX, or 6XX Response

          Q.850 Cause code

          Call Admission Control limits

To configure PCD use the steps below

1.      Define buffers and buffer sizes

          per-call num-buffer
          per-call buffer-size debug

2.      Turn per-call debugging on/off

          per-call shutdown
          per-call active debug
          per-call inactive

3.      Set trigger points

          per-call trigger cause 1
          per-call trigger cause 41
          per-call trigger sip-message 404
          per-call trigger sip-message 488

4.      Export debug buffer
          perc-coalnl etxeponrtt p rimary [flash | ftp | http | pram | rcp
|           tftp] secondary [flash | ftp | http | pram | rcp | tftp]

5.      Show buffer content status
          show per-call stat
          show per-call buffer list

6.      Show buffer contents on console

          router#show per-call buffer content ?
          <0-10000000> Specify the buffer num
          router#show per-call buffer content 1

Configuration example:

per-call num-buffer 10
per-call buffer-size debug 1000
per-call export primary ftp://cisco:cisco@128.107.106.184/PCD1.txt secondary flash0: per-call trigger sip-message 404

As you can see with this feature, CUBE debugs will only be triggered for specific trigger points you are looking at or interested in.

Please rate all useful posts

"opportunity is a haughty goddess who waste no time with those who are unprepared"